Tuesday, March 27, 2012

Density in Los Angeles: a Matter of Tolerance?

Why LA can build densely. Source: Google Streetview
In the previous post, I claimed that building coverage for residential development, based on North American block usage norms, likes to stay under 70% building coverage since building layouts need to grant adequate daylight exposure to dwelling units. As I began to think more about this, I realized that there are notable exceptions, local norms were high building coverage is possible and even tolerated. Some of these exceptional places are the compact subdivisions of Western/Southwestern U.S. and great swaths of urban Los Angeles.  I have actually never heard others remark on the unique qualities that make LA well-suited for high building coverage, but its set of complimentary conditions - including climate, building typologies, cultural history and block morphology - should be appreciated more about LA.  The remarkable fact that a bountiful portion of LA's blocks incorporate relatively high building coverages with multifamily development may be one of the primary reasons why the city has such a high overall population density (average people per square mile) in comparison with other cities. (By the way, over at Discovering Urbanism, Daniel Nairn has just posted a really handy synopsis of the city density trends revealed by the 2010 Census.)

One way to achieve high coverage density is through block morphology.  You can conceivably create a block pattern that accommodates 100% building coverage for multifamily lots. 100% lot coverage is actually possible on the right kind of block: a very, very narrow block of 130-feet wide at the most (measured street centerline to street/alley centerline) so that you can daylight the units on both sides of a building without providing any private open space at all. You just use the rights-of-way to daylight units. (Such a block structure would actually allow Ed Glaeser to claim we can build 150 1600 sq. ft. dwelling units per net acre without having to go higher than six stories.)  But in the prewar urban fabrics, block widths rarely consistently go under 330 feet in most cities.  Of all the North American city grids that I know, only historic Savannah has a block width as narrow as 130 feet as a normative feature of its grid (these are the "Trust Lot" blocks that face the east and west sides of each historic district square). But a 200-foot wide block, as in Manhattan and Portland, does allow these cities to get high coverages, because relatively little open space can be provided in the interior of the blocks when you have 100-foot parcel depths.

A subdivision in Las Vegas, Nevada, demonstrating the 100-foot deep lot morphology that is pervasive there. (Source: Google Maps
In the West and Southwest, subdivision blocks are also commonly this narrow, allowing for some of the highest building coverages you can find in single-family neighborhoods.  These lots grant homeowners some of the tiniest backyards in America.  In time, as in Savannah (whose town lots are also this deep), these 100-foot deep parcels can conceivably redevelop into denser Savannah-style townhouses, apartments, offices and mixed use buildings. One might argue that it may take longer than the history of Savannah to achieve Savannah's urbanism in these subdivisions, perhaps, since their patterns are conceived mainly to keep strangers out, but the hope to realize a modicum of urbanism is there theoretically.  In these cases, one may achieve Jane Jacobs's lower threshold for urban vitality (in the ballpark of 100 dwelling units per net acre) without, perhaps, needing to build above 6 stories. The narrow lot just makes the overall utilization of land more efficient. That assumes however that you transition from single-family development to attached brownstone-style walk-ups or mid-rise development created by assembling rows of single-family lots.  We'll call this strategy for intensification of land usage the "Savannah Strategy": the strategy of assembling and building higher on shallow single-family lots. The morphology of the block pattern is important to this strategy: how narrow the blocks are and how well the streets can be connected in time.

A block interior in Boston's Back Bay. (Source: Google Streetview)
But placing homes tightly together without actually going all the way and attaching them grants an important advantage with respect to average density.  Actually two important advantages (the second of which we will get to later).  We mentioned already exactly what it is: granting access to daylight.  If buildings have to share walls, then they only have the street-facing side and the back-side for daylight access.  That means the building will rarely get deeper than 70 feet.  Boston's Back Bay is an excellent illustration of this fact. See how much open space in the interior of the block can't be used for building coverage...
A typical block in the Back Bay with building perimeters highlighted for clarity. (Source: Google Maps)
However, note also that for Boston's climate this is very good. You do appreciate the adequate daylight that falls into the space, and notice the pleasant balance between shadow and sunlight at midday in the Streetview image above. Note that the open space to building height ratio is 1:1, around what New Urbanists recommend for alleys and pedestrian rights-of-way. At the ends of the blocks, taller buildings at the corners of the block seal up the sides of the block, creating an intimate semi-private environment in the block interior. You can maybe claim that the Back Bay has the perfect block typology for Boston's climate and the building coverage seems appropriate.  

Now compare the Back Bay with this plan view of multifamily buildings in LA...


Both maps are at the same scale.  If Jane Jacobs observed that building coverage over 70% is "intolerable" in Boston's North End, LA doesn't seem to care. As with the shade-loving Western subdivision, this is where LA's sunshine-soaked climate grants LA one dramatic advantage. In LA, people like shady courts. Snug closeness between neighboring buildings (instead of party walls) is tolerated because direct sunlight is just not as coveted and the fact that there is just an accrued cultural tolerance for detached nearness. This means the buildings can be very deep indeed from front to back. The entire building perimeter serves that all-important function of daylighting units and, therefore, you can go deep into the block with many units at level, with units between the front side and back side units.

In sum, high density in LA is achieved with lot coverage, not building height. To make the comparison visually clear, here's a block to block comparison between the Back Bay and Central LA blocks at the same scale...


A typical block in the Back Bay with building perimeters highlighted for clarity. (Source: Google Maps)

A typical block in Central LA. Note the abundant perimeter available for daylighting units. (Source: Google Maps).
The interesting thing about LA is that LA's building typologies, a product of its cultural legacy, free it from the need to achieve high lot coverages with the "Savannah Strategy", i.e. using lots and lots of right-of-way land to daylight units.  In fact, in LA, the fatter the block the better. Part of the reason the courtyard apartment typology suits it so well is that the half-blocks are just deep enough (in LA, the 330-foot wide streetcar suburb block pattern predominates). This is remarkable for one simple reason: it cuts down on the amount of right-of-way land needed overall.  LA needs fewer streets! And so, it bumps up average density thus across the city. Of course, from the standpoint of Jane Jacobs, this quality of LA poses liabilities.  But it does suggest interesting ways to begin to work in LA (and other sun-blessed cities in California and the Southwest) to achieve urbanist goals with qualities other cities simply don't possess in great abundance.  We can intensify land uses by encouraging more mixture of primary uses in the multifamily fabric, allowing some parcels to go high, and connecting the city better with high frequency transit and bike and pedestrian supporting urban design to lessen dependence on automobile storage. We'll call this the "LA Tolerance Strategy": a strategy where block morphology is just not as important as a cultural tolerance for alternative means.

Sunday, March 11, 2012

Squeezing Jane Jacobs into Mid-rise Urbanism


Hull Street, Boston, originally uploaded by Flickr user Asten.

It seems whenever urbanists discuss development and density issues, I often encounter an assumption that travels widely among urbanists: that the urbanism of Jane Jacobs looks no higher than the mid-rise building environment seen in the photo above. No doubt that the urban village Jane Jacobs loved in The Death and Life of Great American Cities is that mid-rise level urbanism of Greenwich Village and Boston's North End (photographed above). Clearly, Jacobs takes in the book what seems to be a kind of pragmatic, middle-ground position on the topic of residential concentration for urban vitality. If density is too low, she argued, it would fail to foster supportive diversity for primary uses, but if it went too high, it would risk imposing standardization of development with building mono-typologies and thus lose the housing variety needed to support a diverse enough population.

The problem here is the numbers.  Jacobs provided a very rough (and qualified) lower and upper limits for "optimal density" in the "Need for Concentration" chapter (Ch. 11), giving the range as 100 to 200 dwelling units per net residential acre. Urbanists, I'm afraid, widely assume the North End is the upper range for this kind of optimized typology and peg her as a mid-rise urbanist from thereon out, never bothering to take a careful look at the numbers provided. Death and Life here is taken wholesale as a paean for 4-6 story urbanity - that "just right" Mama Bear density that neither creates shadowy downtown canyons, nor 2-level rowhouse sprawl. Frankly, I suspect many urbanists interpret Jane Jacobs's requirement that "most blocks be short", her second condition for generating diversity for urbanity, as meaning staying short height-wise (she meant length-wise).

In this 4-6 story vision of neo-Jacobsian urbanism, we urbanists are quite comfortable operating, I must say. Our personal experience of Greenwich Village and the North End bears it out. We jibe with that kind of close-fitting, yet not too high, urbanity. If buildings got any higher, our snug alleys and pocket parks become subsumed in ominous Gotham city shadows.

But as a land planner, this chapter always startles me when I see the numbers.  In no way do I associate, as Ed Glaeser does, 6 stories with 150 dwelling units per net residential acre! This is at least 8 stories, and that's if parking areas/structures were not to count as part of the residential acreage.  With parking,150 dwelling units per net residential acre (150 DUA for short) looks like a district with a healthy mixture of development that includes many 10+ story high rises. Think Portland's Pearl District.

Pearl District, Portland, Oregon
Why 150 DUA needs buildings taller than 6 stories is a matter of how buildings use blocks.  Most importantly, residential units just do not like to get any deeper than 35 feet from the nearest window.   This fact alone means that your footprint can't match the lot perimeter.  When your blocks are 330 to 400-foot wide (as they are for the vast majority of pre-war urban districts in North America), mid-rise multifamily lots are pretty much stuck below 70% building coverage, not just for architectural taste, but for human needs. (In Portland, the smaller blocks help bump up coverage efficiencies. ...As Jacobs said, frequent streets are good!)

Yes, architects can cut into the open space with projecting wings (like those skinny Brooklyn flats with the plans that look like the white piano keys), but, what that means - and why we are loathe to do so beyond 70% coverage - is that your lower units start losing access to sunlight at 4+ story heights. Even Jacobs notes the disadvantage of coverage that is too high in Chapter 11, discussing the North End in particular. The block on the left side of the North End photo at the top, for example, had 72% building coverage in 1960 - way too high for comfort for her (actually, she called it "intolerable"). That is why it was 123 DUA in that 4-7 story height range.

For most 1-5 acre lots, 60% lot coverage for 6+ story multifamily development is a sober number not to surpass for your development. If you want to hit 150 DUA with this, you will need to go to at least 8 stories to secure adequately sized multi-bedroom units.  That's what 150 DUA looks like at a bare minimum with underground parking.  If we elect to squeeze 150 DUA into 6 stories instead, we are going to be building too many one bedroom units - exactly what we shouldn't be building, according to Jacobs, if we want to promote diversity! To use her terms, that would "standardize" your development to stamp out diversity.

In other words, the six stories Ed Glaeser allotted to Jacobsian urbanism is actually exactly what will undermine it by squeezing it from the other side: it will either not create enough density or not enough diverse housing with density. He would be making his point much more pungently if he enlisted Jane Jacobs as an ally in his argument for going taller. 

More than likely, you are going to be using a lot of high-rises to get anywhere near 150 DUA district-wide.  Especially if you are going to be building a diversity of housing products.

Part of adding diversity is building affordably, and for that you also need construction efficiencies that make it worthwhile for developers. Because the price of steel is so high compared to wood-frame construction (which can't get higher than 6 stories), developers don't like to use it in that vaunted luxury mid-range of 7-12 stories, unless they are building in a market that actually can support that product. I would argue that, in fact, we have to climb higher, up to 170 DUA at least, in order to get enough supportable high-rises that can add affordable family unit products to a district. We need to escape the middle!

Moreover, this makes good sense for architectural reasons. The mixed-use buildings in the 4-6 level range are needed to provide "relief" for daylighting their taller brethren. That mixed typology in a district to me seems what we should be extolling as "neo-Jacobian urbanists" (the Vancouver strategy). By easing regulatory pressures to going higher, urbanist development is able to become more affordable again to the middle class, and I can only see Jacobs applauding Glaeser here. Where she would disagree is that we can settle on one solution to suit every frame; in fact, the mixtures and exceptions are important. Sometimes, regulation promotes. As an architect who has to think about things like daylight and the needs of humans and sheer construction realities, i.e. "regulations" of a sort, I have to add that the physical mixture of diverse architectural products, short and tall, side by side, can secure multiple benefits.

In short, don't trust Ed Glaeser when he talks about architecture. As Adam Christian states, "Glaeser conceives of cities first and foremost as consisting of people and connections, and secondarily of places and buildings." Indeed. (Only the cool imagination deduces six-story walk-ups for one hundred fifty 1600 sq.ft. apartments in an acre!) But, urbanists, we gotta take Ed's advice just the same, for his and Jane's reasons, ...and go high. Very high.